Wave model
Andrew Hsiu
January 2019 working draft
Waves of expansions
Map: https://drive.google.com/open?id=19mUt6CE7K7z4T9hzQ_7XAQKjrWyapb-G&usp=sharing
The Sino-Tibetan expansion can be (better) thought of as multiple waves of concentric circles expanding out from the Upper Yangtze region. Trees only tell part of the story; the wave model (and my newly proposed "pyramid model" depicting linguistic strata) tells another side of the story. Traditionally, linguists prefer neatly branching trees, but as Matisoff (2013) noted in his paper on Asakian classification, this is a very crude method that cannot fully explain Sino-Tibetan diversification processes. Archaeobotanical studies have also noted this pattern of multiple waves of movements from the Upper Yangtze region into Northeast India.
Concentric waves (Wikimedia Commons):
For Eastern Sino-Tibetan languages:
Wave 1 (early Neolithic; most basal / aberrant branches, all with non-ST substrata):
Koro
Idu-Taraon
Hrusish
Kho-Bwa
Tujia
Sinitic (earliest layer)
Wave 2 (Middle to Late Neolithic expansions from the Upper Yangtze):
Nungish
Karenic
Gong (expansion southward starting from Yunnan)
pre-Kathu
Sinitic (secondary layer)
"Donor Jiamao"
"Donor Kra"
"Donor Hmong-Mien"
Wave 3 (pre-Burmo-Qiangic expansions from the Sichuan Basin and northern Yunnan; the branches below are the more divergent branches of the Burmo-Qiangic convergence area, which is due to their having split off before the Burmo-Qiangic linkage started to exist):
SE Chamdo
rGyalrong
Horpa-Lavrung
pre-Ersuic
Wave 4 (Bronze Age; Burmo-Qiangic convergence and expansion):
(Core) Qiangic
Lolo-Burmese
Naic
For Western Sino-Tibetan languages:
Wave 1 (most basal / aberrant Western ST branches; both before and during the spread of Austroasiatic into NE India ~3,500-4,000 years ago; links are ultimately with pre-Burmo-Qiangic branches [rGyalrong, Horpa-Lavrung, Ersuic, Lolo-Burmese, etc.] and Nungish in the east, similar to how Malayo-Polynesian languages have links with multiple Formosan branches):
Miju, Meyor
Tani
Lepcha
Mru-Hkongso (prior to Central ST convergence)
pre-Sal (prior to Central ST convergence)
Gongduk
Ole
Wave 2 (subsequent spread of Western ST branches into Nepal, etc.):
West Himalayish
Raji-Raute
Greater Magaric (Kham, Magar, Chepang; perhaps also Dura)
Newaric
Kiranti
Lhokpu-Dhimalish
pre-Tshangla
Wave 3 (secondary expansions during the Late Bronze Age; a Central ST convergence area starts to appear):
Kuki-Chin-Naga
Pyu
Sal
etc.
Wave 4 (Iron Age expansion of Bodish, which started out as Western ST but then mixed with various non-Bodish languages):
Basum
Tamangic
Kaike
East Bodish
Tibetic
The following branches contain non-Sino-Tibetan substrata along with superstrata of early Sino-Tibetan splits. A similar parallel would be the shift of Negritos in the Philippines from non-Austronesian languages to early forms of Malayo-Polynesian (see Reid 2013).
Lepcha: AA > early WST
Mru-Hkongso: AA > early CST
Koro: Siangic > early ST
Idu-Taraon: Siangic > early ST
Hrusish: pre-Hrusish > early ST
Kho-Bwa: pre-Kho-Bwa > early ST
Tujia: pre-Tujia > early ST
Sinitic: pre-Sinitic > early ST
Kuki-Chin-Naga, Sal, Tani, Karenic, and Lolo-Burmese also have many Austroasiatic loanwords, but not enough influence to actually have Austroasiatic substrata.
Maps
1st wave of ST expansion (early EST)
2nd waves of ST expansion (WST, later EST)
3rd waves of ST expansion (Western / later WST, pre-Burmo-Qiangic)
4th waves of ST expansion (CST, Burmo-Qiangic)
5tth wave of ST expansion (Bodish)